:i,VSSKrSSKJr SSKJr SSKJrJr SSK J r "SS \ 5r g) N)links)Model) MaskedModelpartition_tree_shuffle) Explainerc^\rSrSrSr\R SSS4U4SjjrSSSSSSS .U4S jjrS r SS jr S r Sr U=r $)PermutationExplainer aThis method approximates the Shapley values by iterating through permutations of the inputs. This is a model agnostic explainer that guarantees local accuracy (additivity) by iterating completely through an entire permutation of the features in both forward and reverse directions (antithetic sampling). If we do this once, then we get the exact SHAP values for models with up to second order interaction effects. We can iterate this many times over many random permutations to get better SHAP value estimates for models with higher order interactions. This sequential ordering formulation also allows for easy reuse of model evaluations and the ability to efficiently avoid evaluating the model when the background values for a feature are the same as the current input value. We can also account for hierarchical data structures with partition trees, something not currently implemented for KernalExplainer or SamplingExplainer. NTc >[RRU5 Uc [S5e[T U]XX5US9 [ UR[5(d[UR5Ul[U5S:a|"SSUR5nURRRURl Xl UR5HupXRRU 'M gg)a'Build an explainers.Permutation object for the given model using the given masker object. Parameters ---------- model : function A callable python object that executes the model given a set of input data samples. masker : function or numpy.array or pandas.DataFrame A callable python object used to "mask" out hidden features of the form ``masker(binary_mask, x)``. It takes a single input sample and a binary mask and returns a matrix of masked samples. These masked samples are evaluated using the model function and the outputs are then averaged. As a shortcut for the standard masking using by SHAP you can pass a background data matrix instead of a function and that matrix will be used for masking. To use a clustering game structure you can pass a ``shap.maskers.Tabular(data, clustering="correlation")`` object. seed: None or int Seed for reproducibility **call_args : valid argument to the __call__ method These arguments are saved and passed to the __call__ method as the new default values for these arguments. Nzmasker cannot be None.)linklinearize_link feature_namesrc:^\rSrSrSSSSSSS.U4SjjrSrU=r$) ;PermutationExplainer.__init__..PermutationExplainer>FautoN max_evals main_effects error_bounds batch_sizeoutputssilentc .>[TU]"UUUUUUUS.6$)Nrsuper__call__ selfrrrrrrargs __class__s _/srv/projetos/modelo_ml_acdoc/venv/lib/python3.13/site-packages/shap/explainers/_permutation.pyr DPermutationExplainer.__init__..PermutationExplainer.__call__?s,!7+"+%1%1#- '%)__name__ __module__ __qualname____firstlineno__r __static_attributes__ __classcell__r$s@r%r r>s!"!&!&%  r'r )nprandomseed ValueErrorr__init__ isinstancemodelrlenr$r __doc__items__kwdefaults__) r"r6maskerrrrr2 call_argsr kvr$s r%r4PermutationExplainer.__init__s4 t >56 6 T`mn$**e,,tzz*DJ y>A  t~~ *59NN4K4K4S4S ) ) 11N!)23 ,,Q/*3 r'rFrrc .>[TU]"UUUUUUUS.6$)z7Explain the output of the model on the given arguments.rrr!s r%r PermutationExplainer.__call__Ys,w %%!  r'c x[URURURUR/UQ76nUS:XaS[ U5-nSn [ URSS5b[URR[R5(aURRn OI[URR5(aURR"U6n O [S5eUR5n [R"[ U5[S9n SX'[R"S[ U 5-S -[ S9n [R"U S 'US[ U 5-S --n SnSnS nSn[ U 5S :Ga[%U 5GHanU b ['XU 5 O[R(R+U 5 S nU H nUU U'US - nM U H nUU U'US - nM U"U S US 9nUck[R"[ U54UR,S S-5nU(a4[R"SU -[ U54UR,S S-5nS nU H<nUU==UUS -UU- - ss'U(aUUS -UU- UUU'US - nM> US - nU H<nUU==UUUUS -- - ss'U(aUUUUS -- UUU'US - nM> US - nGMd U S :Xa![/S US S[ U 5-S -S35eUS nU(aUR1XS9nO[R"S [ S9n U"U S S S 9nUS n[R"[ U54UR,S S-5nU(a4[R"SU -[ U54UR,S S-5nUSU -- UUR2UU UcSOUR5S 5[7URS5(aURR8S.$SS.$)z_Explains a single row and returns the tuple (row_values, row_expected_values, row_mask_shapes).rN clusteringzeThe masker passed has a .clustering attribute that is not yet supported by the Permutation explainer!)dtypeTrrr) zero_indexrz max_evals=zV is too low for the Permutation explainer, it must be at least 2 * num_features + 1 = !)r output_names)valuesexpected_values mask_shapesrrD error_stdrH)rr6r;rrr7getattrr5rDr0ndarraycallableNotImplementedErrorvarying_inputszerosboolintdelta_mask_noop_valuerangerr1shuffleshaper3rrKstdhasattrrH)r"rrrrrrrow_argsfmrow_clusteringinds inds_maskmasks npermutations row_valuesrow_values_history history_posmain_effect_values_iindexpected_values r% explain_row PermutationExplainer.explain_rownsT[[$))T=P=P \S[ \  R(I 4;; d 3 ?$++00"**==!%!7!7$++0011!%!7!7!B){   "HHSWD1  SY*#644a!a#d)ma&78  ! ! t9q==)!-+4NKII%%d+C"E!HFA  C"E!HFA UqZH%!#3r7*w}}QR7H*H!IJ#-/XX !M 1 #B&mmAB/ 0.*CsOwq1u~ 'BBO#?Fq1u~PWXYPZ?Z*;7<FA  q CsOwqzGAEN'BBO#?FqzGTUXYTYN?Z*;7<FA  q ]*`!   ,BCDGJKOGPCPSTCTBUUVW%QZN%'__T_%Q"HHQc*E1;G$QZN3r7*w}}QR/@"@AJ%'XXM)BmmAB' (&"!A $56->>.(!3!;ASAWAWXYAZ7>tzz>7Z7ZDJJ33  ae  r'cHU"XURS-US9nUR$)aYLegacy interface to estimate the SHAP values for a set of samples. Parameters ---------- X : numpy.array or pandas.DataFrame or any scipy.sparse matrix A matrix of samples (# samples x # features) on which to explain the model's output. npermutations : int Number of times to cycle through all the features, re-evaluating the model at each step. Each cycle evaluates the model function 2 * (# features + 1) times on a data matrix of (# background data samples) rows. An exception to this is when PermutationExplainer can avoid evaluating the model because a feature's value is the same in X and the background dataset (which is common for example with sparse features). Returns ------- array or list For models with a single output this returns a matrix of SHAP values (# samples x # features). Each row sums to the difference between the model output for that sample and the expected value of the model output (which is stored as expected_value attribute of the explainer). For models with vector outputs this returns a list of such matrices, one for each output. r)rr)rXrI)r"Xrarr batch_evalsr explanations r% shap_values PermutationExplainer.shap_valuess*21 (BQ]^ !!!r'cg)Nz&shap.explainers.PermutationExplainer()r()r"s r%__str__PermutationExplainer.__str__s7r')r$r6) FFTF)r)r*r+r,r8ridentityr4r rjrprsr-r.r/s@r%r r sT #(..UY`d@4L  *q f"888r'r ) numpyr0rmodelsrutilsrr _explainerr r r(r'r%r|s"7!u89u8r'